Abuse liability and reinforcing efficacy of oral tramadol in humans

Abstract

Background: Tramadol, a monoaminergic reuptake inhibitor, is hepatically metabolized to an opioid agonist (M1). This atypical analgesic is generally considered to have limited abuse liability. Recent reports of its abuse have increased in the U.S., leading to more stringent regulation in some states, but not nationally. The purpose of this study was to examine the relative abuse liability and reinforcing efficacy of tramadol in comparison to a high (oxycodone) and low efficacy (codeine) opioid agonist.

Methods: Nine healthy, non-dependent prescription opioid abusers (6 male and 3 female) participated in this within-subject, randomized, double blind, placebo-controlled study. Participants completed 14 paired sessions (7 sample and 7 self-administration). During each sample session, an oral dose of tramadol (200 and 400 mg), oxycodone (20 and 40 mg), codeine (100 and 200 mg) or placebo was administered, and a full array of abuse liability measures was collected. During self-administration sessions, volunteers were given the opportunity to work (via progressive ratio) for the sample dose or money.

Results: All active doses were self-administered; placebo engendered no responding. The high doses of tramadol and oxycodone were readily self-administered (70%, 59% of available drug, respectively); lower doses and both codeine doses maintained intermediate levels of drug taking. All three drugs dose-dependently increased measures indicative of abuse liability, relative to placebo; however, the magnitude and time course of these and other pharmacodynamic effects varied qualitatively across drugs.

Conclusions: This study demonstrates that, like other mu opioids, higher doses of tramadol function as reinforcers in opioid abusers, providing new empirical data for regulatory evaluation.

Figure 1

Mean pupil diameter (top panel) and expired CO₂ (bottom panel) after administration of oxycodone (left column), tramadol (middle column) and codeine (right column) as a function of time following drug administration across the 6-hour session (n=9, ±1 SEM). Time course analysis revealed a significant effect of dose on pupil diameter (F[6,48]=14.79, p<.001), and Dunnett post-hoc tests indicated both doses of oxycodone and the high doses of tramadol and codeine decreased pupil diameter relative to placebo (p<.05). A main effect of dose was also detected on expired CO₂ concentrations (F[6,48]=7.98, p<.001), with a significant effect of both oxycodone and tramadol doses and 200 mg codeine on expired CO₂, relative to placebo (Dunnett post-hoc, p<.05). Filled symbols indicate means that were significantly different from placebo at a particular time point (Dunnett post-hoc, p<.05).

Figure 2

Mean VAS ratings of the subjective measures Drug Effect (top panel), Like Drug Effect (middle panel) and Bad Drug Effect (bottom panel) after administration of oxycodone (left column), tramadol (middle column) and codeine (right column) as a function of time following drug administration across the 6-hour session (n=9, ±1 SEM). Time course analysis indicated a main effect of dose on VAS Drug Effect (F[6,48]=5.6, p<.001), Like Drug Effect (F[6,48]=4.8, p<.001) and Bad Drug Effect (F[6,48]=3.8, p<.004). Dunnett post-hoc tests indicated both doses of oxycodone increased ratings of Drug Effect and Like Drug Effect, while the high dose of tramadol increased ratings of Bad Drug Effect, relative to placebo (p<.05). Filled symbols indicate means that were significantly different from placebo at a particular time point (Dunnett post-hoc, p<.05).

Figure 3

Mean number of trials completed for drug as a function of dose condition, with small circles representing the number of trials completed by individual subjects (n=9, ±1 SEM). A main effect of dose (F[6,48]=5.7, p<.001) was detected and Dunnett post-hoc tests indicated all doses except the low dose of tramadol were significantly different from placebo (p <.05).

Figure 4

Mean breakpoints for drug and money as a function of dose condition, with filled squares representing mean drug breakpoints and open circles representing mean money breakpoints (n=9, ±1 SEM). A main effect of dose (F[6,48]=5.2, p<.001) was detected on drug breakpoints and all breakpoints except those under the low dose of tramadol and the high dose of codeine were significantly different from placebo (Dunnett post-hoc, p <.05). A main effect of dose was also detected on money breakpoints (F[6,48]=4.8, p<.001) and all money breakpoints were significantly different from placebo (Dunnett post-hoc, p<.05).